Pretrained deep-learning models are the go-to solution for images or text. However, for tabular data the standard is still to train tree-based models. Indeed, transfer learning on tables hits the challenge of data integration: finding correspondences, correspondences in the entries (entity matching) where different words may denote the same entity, correspondences across columns (schema matching), which may come in different orders, names... We propose a neural architecture that does not need such correspondences. As a result, we can pretrain it on background data that has not been matched. The architecture –CARTE for Context Aware Representation of Table Entries– uses a graph representation of tabular (or relational) data to process tables with different columns, string embedding of entries and columns names to model an open vocabulary, and a graph-attentional network to contextualize entries with column names and neighboring entries. An extensive benchmark shows that CARTE facilitates learning, outperforming a solid set of baselines including the best tree-based models. CARTE also enables joint learning across tables with unmatched columns, enhancing a small table with bigger ones. CARTE opens the door to large pretrained models for tabular data.

One of the interesting characteristics of crime data is that criminal cases are often interrelated. Criminal acts may be similar, and similar incidents may occur consecutively by the same offender or by the same criminal group. Among many machine learning algorithms, network‐based approaches are well‐suited to reflect these associative characteristics. Applying machine learning to criminal networks composed of cases and their associates can predict potential suspects. This narrows the scope of an investigation, saving time and cost. However, inference from criminal networks is not straightforward as it requires being able to process complex information entangled with case‐to‐case, person‐to‐person, and case‐to‐person connections. Besides, being useful at a crime scene requires urgency. However, predictions from network‐based machine learning algorithms are generally slow when the data is large and complex in structure. These limitations are an immediate barrier to any practical use of the criminal network geared by machine learning. In this study, we propose a criminal network‐based suspect prediction framework. The network we designed has a unique structure, such as a sandwich panel, in which one side is a network of crime cases and the other side is a network of people such as victims, criminals, and witnesses. The two networks are connected by relationships between the case and the persons involved in the case. The proposed method is then further developed into a fast inference algorithm for large‐scale criminal networks. Experiments on benchmark data showed that the fast inference algorithm significantly reduced execution time while still being competitive in performance comparisons of the original algorithm and other existing approaches. Based on actual crime data provided by the Korean National Police, several examples of how the proposed method is applied are shown.

Most of the content-based image retrieval systems require a distance computation for each candidate image in the database. As a brute-force approach, the exhaustive search can be employed for this computation. However, this exhaustive search is time-consuming and limits the usefulness of such systems. Thus, there is a growing demand for a fast algorithm which provides the same retrieval results as the exhaustive search. We propose a fast search algorithm based on a multiresolution data structure. The proposed algorithm computes the lower bound of distance at each level and compares it with the latest minimum distance, starting from the low-resolution level. Once it is larger than the latest minimum distance, we can remove the candidates without calculating the full-resolution distance. By doing this, we can dramatically reduce the total computational complexity. It is noticeable that the proposed fast algorithm provides not only the same retrieval results as the exhaustive search, but also a faster searching ability than existing fast algorithms. For additional performance improvement, we can easily combine the proposed algorithm with existing tree-based algorithms. The algorithm can also be used for the fast matching of various features such as luminance histograms, edge histograms, and local binary partition textures.

A traditional method for down conversion of MPEG compressed video performs low-pass filtering and subsampling after full decompression. However, the method has two shortcomings: large memory requirement and high computational complexity. To overcome the drawbacks, recent research has focussed on down conversion in the DCT domain. Existing DCT domain methods need only a quarter of full frame memory because completely reduced frames are stored in the frame memory. But the loss of field information often causes significant error during motion compensation. To solve this problem, we adopt a hybrid structure where field information is preserved by performing down conversion in the DCT domain only for horizontal direction and down conversion in the spatial domain for vertical direction. Furthermore, we present a novel technique for the horizontal down conversion which adopts a modified 4-IDCT kernel. The modified 4-IDCT kernel outperforms the conventional 4-IDCT kernel in that it performs low-pass filtering and sub-sampling additionally. Experimental results show that the proposed scheme provides high visual quality while maintaining comparable complexity.